© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW
    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    2. Gemma-2-27B
    3. 10-GEMMASCOPE-RES-131K
    4. 3815
    Prev
    Next
    INDEX
    Explanations

    verbs followed by articles or exceptions

    np_acts-logits-general · gemini-2.5-flash-lite

    This neuron spotlights the key question verb (auxiliary or main verb) in headlines—e.g. "Get," "become," "be," "can," "mean"—that frames the core of a query.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_10/width_131k
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     and
    -3.89
     as
    -1.67
     by
    -1.62
    lossians
    -1.57
    ны
    -1.55
     one
    -1.54
     until
    -1.52
    _{\
    -1.47
     with
    -1.41
    )};
    -1.40
    POSITIVE LOGITS
    0
    2.08
     sorprende
    2.02
    2
    1.86
    1
    1.84
    无疑
    1.80
     tradução
    1.73
     gratuita
    1.72
    還會
    1.71
     multiuso
    1.70
     الرسمي
    1.67
    Activations Density 0.060%

    No Known Activations